26 
equivalents of work which may be presented in a very 
diverse and hardly recognisable manner. 
In his first Essays, Art. XXXIX., “On an Absolute 
Thermometric Scale,” and Art. XLI., “An Account of 
Carnot’s Theory of the Motive Power of Heat,” dating 
from the years 1848 and 1849, our author still occupies 
essentially Carnot’s standpoint, but he nevertheless calls 
attention to the fact that the argument adduced by Carnot 
in support of his theorem, apparently valid though it was 
in all points, was yet defective if the experiments by 
Joule, which were just then made known, should be con- 
firmed, according to which heat might be generated anew 
by work (vol. i. p. 116). That which more immediately 
directed Sir William Thomson’s studies to this subject 
was the possibility of attaining, in accordance with 
Carnot’s theorem, to an absolute scale of temperature, 
and he endeavoured to utilise the observations which 
Regnault had shortly before carried out with special care 
in reference to the pressure and latent heat of steam for 
he purpose of calculating such a scale. But in doing so, 
he was obliged to apply the hypothesis, not perfectly exact 
in this case, that the density of steam was to be calcu- 
lated from pressure and temperature according to the 
laws of gases. 
The theory of Carnot next obtained highly surprising 
confirmation from the theoretical deductions drawn by 
Prof. James Thomson, the elder brother of Sir William, 
touching the alteration of the freezing-point of water in 
consequence of differences of pressure. The accuracy in 
point of fact of this deduction was experimentally demon- 
strated by Sir W. Thomson. This was a discovery which 
perhaps more than any other served to draw the attention 
of physical scientists to the accuracy and the importance 
of Carnot’s theorem. 
Meanwhile our author, no longer able to doubt the cor- 
rectness of Robert Mayer and Joule’s thesis respecting 
the equivalence of heat and work, devoted himself to the 
problem of how Joule’s and Carnot’s laws might be 
combined. This question he answered in his treatise of 
March, 1851, “On the Dynamical Theory of Heat,” 
Art. XLVIII. Prof. Clausius,in Germany, had, however, 
been busied with the same problem, and had published 
the results at which he arrived before Sir W. Thomson, 
in May, 1850. ‘The essential results of the two investi- 
gations coincided exactly ; only in their numerical values 
for the absolute scale of temperature, the two authors had 
started with two different hypotheses, and had therefore 
reached different conclusions. Sir William Thomson 
had, as above mentioned, calculated the density of steam 
from pressure and temperature, as if for complete gases, 
whereas Prof. Clausius had accepted the hypothesis set 
up by Robert Mayer, according to which the work of a 
gas expanding itself was exactly equivalent to its loss of 
heat. Later on, when his opponents set forth the un- 
satisfactory basis of this hypothesis, Robert Mayer 
pointed to an old and very little-known experiment of 
Gay-Lussac, according to which a gas diffusing itself in 
empty space without encountering any resistance suffered 
no diminution of heat. The same experiment was after- 
wards carried out by Joule without his having any know- 
ledge of the earlier observation of a similar nature. This 
form of the experiment was, however, as a whole, not 
fitted to yield very precise results, seeing that the mass of 
NATURE 
| Way 14, 1885 
air available for it, whose consumption of heat was to be 
measured, was necessarily very small in comparison with 
the mass of water of the calorimeter. It was not till the 
investigations into the changes of temperature undergone 
by a mass of gas made to pass through a very dense 
porous substance—an investigation carried out in common 
by J. P. Joule and Sir W. Thomson, in 1852, and described 
in Art. XLIX., “On the Thermal Effects of Fluids in 
Motion ”—that it was demonstrated how, in point of fact, 
R. Mayer’s hypothesis was accurate to within a very 
close degree of approximation, although not with abso- 
lute precision, in respect of hydrogen and atmospheric air, 
whereas carbonic acid showed greater deviations. 
To this have to be added extended investigations into 
thermo-electric currents, and the equivalent of their 
operations (Appendix to Art. XLVIII. and Art. LI. 
“Experimental Researches in Thermo-electricity,” Vol. I., 
Art. XCI. Bakerian lecture, pp. i., ii., and iii., Vol. II.). In 
a thermo-electric chain which, from its conducting wire, 
sets magnets in motion, or generates heat in them, the 
heat conducted to the soldering seams is manifestly the 
source of the operations. We know that in such a case, 
according to the important observations of Peltier, heat 
disappears from the warmer soldering seam, and becomes 
developed in the colder. That is, in fact, the condition, 
according to Carnot’s law, under which heat becomes 
transferable into other forms of work. This particular 
process was, however, of special interest for the universal 
validity of the theory, seeing that the work of heat is here 
produced under conditions altogether different from those 
of the steam-engine and hot-air engine. Our author was 
by this investigation led to the conclusion that, contrary to 
the opinion hitherto entertained, it was not in the solder- 
ing-seams of the metals, at all events not in those alone, 
but in the whole length of the wires, by a process which 
he calls “electric convection of heat,” that the essential _ 
cause of the thermo-electric force was to be sought ; and, 
in point of fact, he succeeded by a series of very laborious 
and subtle experiments in demonstrating that the conduc- 
tion of heat in iron proceeded more rapidly in the direction 
of the current of negative electricity, and in copper in the 
direction of the positive current. 
In the first volume of the book which is the subject of 
notice, the consecutive stages may thus be followed in the 
development of one of the most remarkable chapters in the 
history of discoveries,a chapter specially remarkable also as 
an example of how discoveries are arrived at in a manner 
not always rational. The course of this discovery re- 
minds one in some measure of the invention of achrom- 
atic telescopes. Starting with the erroneous supposition 
that the eye of man was achromatic, Euler inferred that 
Newton’s assumption of the proportionality between 
refraction and dispersion of light was false, and that 
his conclusion as to the impossibility of achromatic 
telescopes was without foundation. Thereupon Euler 
gave the receipt for the making of achromatic tele- 
scopes—a correct conclusion from a false premiss; 
similar to the case of Carnot with the doctrine of heat. 
After all the confirmations which have been obtained in 
the different branches of physics for the validity of the 
deductions of the corrected Carnot law there can hardly 
longer remain any doubt that we have here found one of 
the most comprehensive and important laws of nature of 
